NSFW What's up with the buffers in cornish designs?

He has a very specific design philosophy about buffers. I think it’s possible that, back in the day, tone suck from very long cables was a real pita and he certainly solved that problem with output buffers. To some extent I don’t think he’s totally wrong: I have definitely experienced issues with modern built unbuffered pedals, things like the tone changes if you turn a delay (typically buffered) on after a drive which is caused by the drive having high output impedance and tone sucking when the delay is off and then the delay “fixing” that when on. That can cause serious problems for touring musicians, which is definitely the audience (I’m not one of them!).
Tell me more... because I have a new DL4 mk2 and I hate it... Whenever I turn it on it adds an unwanted boost and a low mid thing... I have no clue how to make it stop, and I was so pissed at this brand new, $300 pedal that I threw it off my board and put my 20 year old original DL4 back on instead
 
Tell me more... because I have a new DL4 mk2 and I hate it... Whenever I turn it on it adds an unwanted boost and a low mid thing... I have no clue how to make it stop, and I was so pissed at this brand new, $300 pedal that I threw it off my board and put my 20 year old original DL4 back on instead
The DL4mkII does have options for true-bypass and analog dry-through that might be worth trying
 
I have had similar experiences...

Things I feel a duty to post:

1. I have been to Cornwall and can attest that some of the pasties are amazing. Better than you might expect from the UK.

2. While I have never heard a Cornish pedal I like I hold no animosity towards either Mr Cornish, his family, or the people of Cornwall. The reasons I don't like his pedals are obviously my own stupid fault/lack of stadium gigs and very long cables.

3. While it can be irritating that the latest hip $500 pedal is simply a Timmy with two parts changed, sometimes changing a few parts in a Timmy can have an extraordinary effect on the sound. If someone (I'm looking at me specifically!) can take a Timmy circuit, change it around enough to get a significantly different and worthwhile sound from it, that's ok by me. I'm sure Paul Cochran didn't design Timmy in a vacuum. A Timmy is kinda what happens when you remove the transistor stages from each end of a Tubescreamer, add a bass pot and simplify the tone control. All of which sounds like a good idea to me.

Really, what matters to me is how good and useful something sounds. If it's affordable then all the better. If it's not I don't buy it. Sometimes as folks fascinated about how things work we lose sight of the original purpose - does it sound good and make me want to play more? If it does - great! If someone finds a way to change two parts in a pedal to make it suddenly sound amazingly different and better for you then that's fantastic!

I admit to being sceptical and disappointed by some new pedal when it is shown to be "merely" a modified whatever. But if it sounds better who cares? The whole reason I got into this hobby was to try and find a better sound for me. I couldn't get the sounds I wanted from what was available. And I still don't like any of the Cornish designs I have tried.

Man have things changed!
 
As usual, I am of the strong opinion that TL072 has no business being in a top notch guitar buffer in 2023. It's just too noisy at 18nV/sqrt(Hz) (same as the thermal noise of a 20k resistor). I would pick something with less than half that noise. And of course, it would also have to have very low current noise to do well with high impedance sources, so either JFET input or super beta bipolar (my new favorite). My best suggestions as a guitar buffer: OPA205/206/207, OPA145. Or, if you have all the money and want the quietest possible op amp that can handle high source Z, get an OPA827, or even two of them in parallel if you're nuts (I am, and I've done it).
Looking at the specs sheet of the tl072 I also came to that conclusion when I first started looking into buffers (not that I had the knowledge to really make such an assertive claim to other people - I still don't hahah). What's you're opinion on the TLE2072?

Edit: I double checked the datasheet for the TLE2072 and the noise is 11.6nV/sqrt typical, so it's better, but not quite at the level of the others you mentioned.
 
Last edited:
What's you're opinion on the TLE2072?
TLE2072 looks like an easy drop-in upgrade over TL072, but I find it stupidly expensive for what it is. If you're going to spend 10x the price of a TL072, then might as well get a much better OPA2205. Only downside (for some) is that it's SMD-only. SMD is not hard, though. Just get the right tools and embrace the damn thing, else you'll get trapped in the past.

Why do I insist on sub-10nV/sqrtwhatever input stages? A guitar pickup can easily have a DC resistance of 5k or less (some single coils and parallel wired humbuckers). The thermal noise of that 5k dictates the lowest achievable noise floor, and your op amp should be at least this quiet, or quieter.
 
Honestly, for me, SMD is fine even with moderately good, standard soldering tools. No need for an SMD soldering station, hot air gun or solder paste. As long as you have a relatively fine soldering iron tip available, I don't see why anyone can't give it a shot. Thanks for the info (y)
 
Last edited:
Honestly, for me, SMD is fine even with moderately good, standard soldering tools. No need for an SMD soldering station, hot air gun or solder paste. As long as you have a relatively fine soldering iron tip available, I don't see why anyone can't give it a shot. Thanks for the info (y)
Agreed. That’s how I do it and it works slick.
 
I did get a bunch of TLE2072 chips for a much more reasonable price a few years ago. The OPA2992 seems like the more reasonably priced opamp that meets your 'requirements', JTEX. May I ask, how did you get from the ~5k DC resistance thermal noise to the figure of < 10nv/sqrt?
 
TLE2072 looks like an easy drop-in upgrade over TL072, but I find it stupidly expensive for what it is. If you're going to spend 10x the price of a TL072, then might as well get a much better OPA2205. Only downside (for some) is that it's SMD-only. SMD is not hard, though. Just get the right tools and embrace the damn thing, else you'll get trapped in the past.
If you just have to use through hole parts, I've had excellent results in my bass preamps with LT1113 and LT1169. But the price has more than doubled in the last few years, and I've never tried them with a low-ish voltage single sided power supply. But as you say, best to move along to SMDs, and buying some tweezers and a few $6 soldering iron tips is hardly a big deal, IMO.
 
I did get a bunch of TLE2072 chips for a much more reasonable price a few years ago. The OPA2992 seems like the more reasonably priced opamp that meets your 'requirements', JTEX. May I ask, how did you get from the ~5k DC resistance thermal noise to the figure of < 10nv/sqrt?
OPA2992 is about half the price of a 2205, same noise spec at 1kHz, but much worse below (CMOS process). It draws 10x more power (!), so it never even made it past my parametric search, since I really care about battery life in my designs.

Thermal noise figures came from my trusty Excel speadsheet calculator:

Resistance (ohms)Noise Density (nV/√Hz)Noise Amplitude (dBu) @ 20kHz BWOutput noise after 50dB of added gain
(fairly typical for heavy metal distortion)
1000 (NE5534, DNU as guitar buffer)
4.0-123-73
2000 (~NE5532, DNU as guitar buffer)
5.7-120-70
3000 (OPA2205)
7.0-118-68
5000 (single coil pickup)
9.0-116-66
8000 (vintage humbucker)
11.4-114-64
16000 (hot humbucker, ~TL072)
16.1-111-61
250000 (500k guitar volume pot @ 50%)
63.6-99-49
 
Last edited:
If you just have to use through hole parts, I've had excellent results in my bass preamps with LT1113 and LT1169. But the price has more than doubled in the last few years, and I've never tried them with a low-ish voltage single sided power supply. But as you say, best to move along to SMDs, and buying some tweezers and a few $6 soldering iron tips is hardly a big deal, IMO.
I long ago stopped even looking at anything from Analog Devices/LT due to their outrageous prices. Hard to believe they actually they sell any, unless it goes on a satellite or something.
 
OPA2992 is about half the price of a 2205, same noise spec at 1kHz, but much worse below (CMOS process). It draws 10x more power (!), so it never even made it past my parametric search, since I really care about battery life in my designs.

Thermal noise figures came from my trusty Excel speadsheet calculator:

Resistance (ohms)Noise Density (nV/√Hz)Noise Amplitude (dBu) @ 20kHz BWOutput noise after 50dB of added gain
(fairly typical for heavy metal distortion)
1000 (NE5534, DNU as guitar buffer)
4.0-123-73
2000 (~NE5532, DNU as guitar buffer)
5.7-120-70
3000 (OPA2205)
7.0-118-68
5000 (single coil pickup)
9.0-116-66
8000 (vintage humbucker)
11.4-114-64
16000 (hot humbucker, ~TL072)
16.1-111-61
250000 (500k guitar volume pot @ 50%)
63.6-99-49
I like your process my friend. Very thorough. The noise of a TL072 really is relatively high there. I assume the NEXXXX opamps should not be used as a buffer due to both of their rather low input impedance figures. Where, then, would they be suitable for use in guitar effects? I've seen people mod their old Boss graphic equalisers with NE5532 opamps, I must assume the current draw is some horrendous.

As an aside, I am very impressed by your projects and schematics on your website. Some truly mindboggling stuff on there!
 
I long ago stopped even looking at anything from Analog Devices/LT due to their outrageous prices. Hard to believe they actually they sell any, unless it goes on a satellite or something.
LT used to be very generous with free samples. ;)

Since I mostly just do one-offs for myself, $10-15 a pop for opamps doesn’t really even phase me as long as I get the results I need. I’m aready burning money and even more so time, why pretend to be rational?
 
Last edited:
How about the OPA1678? It seems better than the OPA2205 in almost every way, except in input offset voltage at 2V for OPA1678 and 0.25V for the OPA2205 and a considerably higher quiescent current for the OPA1678. If you don't care about that much about the power consumption for a non battery powered device, the considerably lower cost of the OPA1678 is very attractive. Or is the reduced input offset voltage of the OPA2205 really worth it?
 
How about the OPA1678?
Looking very good. It was off my radar due to my low power budget, but if that's not a concern, then at first glance it seems to be a terrific op amp, especially when considering the price. It's not even that power-hungry. Input offset is not a factor for our audio purposes here. Great find!
 
The DL4mkII does have options for true-bypass and analog dry-through that might be worth trying
I have heard this... I tried to consult the "manual" (pdf only), but it is kinda tech-speak. I admit, I'm a dumb hillbilly, and when it comes to "stop playing, hold this button, and while holding this button, push this combination til this light flashes, and now you will be in such and such mode... from there, release this button while pressing this other button until you get three flashes, then hurry and turn this knob... blah blah".

Meanwhile you accidentally hit the wrong thing and now everything is all screwed up and you have to do a factory reset and try to get back to square one... Just way too complicated... Then you try to get tech support and "hook the device up to this software and download the update"...

I have to dig out my laptop that magically went obsolete last year and now the software is no longer supported...

Gah, I hate technology...
 
Motivated by the new aion traces, a question has emerged.

If I'm not horribly mistaken (wouldn't be the first time), the buffers in those Cornish designs are bootstrapped emitter followers. I know this allows for higher input impedances, but is it worth the extra effort for a guitar pedal?

Thanks!

Violating my own code here by bumping what became a sort of controversial thread but what the hell…

I think the main draw here, if I can presume to try and get inside Mr. Cornish’s thought process, is that he is already using the BC549C in a bunch of pedals.

One of the reasons I’m bumping this is because I’ve spent the last few days revising the excel calculator I used for gain stages and buffers. One thing I really wanted to understand, and see for myself, was how the design parameters can affect input and output impedance.

Once I had my calculator revised and check against multiple known calculators, and the having run the calculations a number of ways, I set up the standard Cornish buffer parameters.

Here is the C-Buffer schematic offered by @Robert :

IMG_2311.jpeg

I ran it with R1, R7, R8 and C5 removed, which is the basic bootstrapped CC topology.

One thing that immediately jumps out at me is how HIGH the input impedance is. With his standard values, the input impedance is something like 4M ohm if I’m not making any mistakes.

When we add back R1, in parallel essentially, you drop to about 2.8M ohm, assuming that a BC549C is around 530 hFE and around 0.65 Vbe, which the majority of mine measure at.

Now, that could be down to what my TC1 is testing at as well, which is always somewhat of an X factor (mental note).

Since I was confident in the calculations themselves, I tried some different values of R3, R4, R5 and R6.

It seems Cornish has it setup so that Ve is about 4.8v.

It’s really difficult to match that when you start changing those resistor values.

The biggest obstacle for me was trying to arrive at some more standard values for SMD resistors.

I think I landed on R3/R5 at 150k and R4 at 220k. Then I dropped R6 to 6.8k and used a 1.5M for R1.

I ended up with around 1M inout impedance and around the same output impedance as the Cornish (~ 50 ohm) without R8.

Going to play a bit with some MPSA13 in here as well, which might give me more flexibility with respect to values.

I was doing this so I can use a modified version as inout and output buffers for an overdrive I’m working on as well as an output buffer for a modified MOSFET Driver in making.
 
Violating my own code here by bumping what became a sort of controversial thread but what the hell…

I think the main draw here, if I can presume to try and get inside Mr. Cornish’s thought process, is that he is already using the BC549C in a bunch of pedals.
It was never my intention to have this thread become what it became when I originally posted it, so I really appreciate you revisiting it with a technical discussion, and what I think is a super reasonable insight
 
It was never my intention to have this thread become what it became when I originally posted it, so I really appreciate you revisiting it with a technical discussion, and what I think is a super reasonable insight

No sweat man. It was fresh in my mind because I was working on it today and this seemed the most logical place to post it.

What I actually found interesting were a few things most of us would see as “kludges” or “fudging it”.

R1 and R8 come to mind. Obviously Cornish was targeting a specific input impedance he couldn't get with the bootstrapped emitter follower alone.

It seems that emitter voltage around 50% of source voltage was important to him, so R1 acting as an artificial reducer of input impedance made more sense than trying to get blood from a stone.

Furthermore, it seems that 100 ohm output impedance was a design criteria so he simple added R8 in series to get there.

If anything, I appreciate him a bit more because of these little practical things.
 
Last edited:
Back
Top